Re: [patch] fsfs revprop packing paranoia

Committed r987512, thanks. Daniel Shahaf wrote on Fri, Aug 13, 2010 at 13:34:55 +0300: > Philip Martin wrote on Fri, Aug 13, 2010 at 11:13:10 +0100: > > Daniel Shahaf writes: > > > > > I figure this patch can't do any harm (except cost another file-read when > > > a > > > write-lock or txn-curr

Re: [patch] fsfs revprop packing paranoia

Philip Martin wrote on Fri, Aug 13, 2010 at 11:13:10 +0100: > Daniel Shahaf writes: > > > I figure this patch can't do any harm (except cost another file-read when a > > write-lock or txn-current-lock is being acquired), and it might as well > > help, so I committed it in r984990. > > >> @@ -594

Re: [patch] fsfs revprop packing paranoia

Daniel Shahaf writes: > I figure this patch can't do any harm (except cost another file-read when a > write-lock or txn-current-lock is being acquired), and it might as well > help, so I committed it in r984990. >> @@ -594,7 +598,11 @@ static svn_error_t * >>err = get_lock_on_filesystem(lock

Re: [patch] fsfs revprop packing paranoia

I figure this patch can't do any harm (except cost another file-read when a write-lock or txn-current-lock is being acquired), and it might as well help, so I committed it in r984990. Daniel Shahaf wrote on Tue, Aug 03, 2010 at 18:59:33 +0300: > Daniel Shahaf wrote on Mon, Aug 02, 2010 at 23:44:10

Re: [patch] fsfs revprop packing paranoia

Daniel Shahaf wrote on Mon, Aug 02, 2010 at 23:44:10 +0300: > Is this necessary to avoid some race conditions around a revprop becoming > packed just before commit_body()'s write-lock had been acquired? > > [[[ > Index: fs_fs.c > ===

[patch] fsfs revprop packing paranoia

Is this necessary to avoid some race conditions around a revprop becoming packed just before commit_body()'s write-lock had been acquired? [[[ Index: fs_fs.c === --- fs_fs.c (revision 981659) +++ fs_fs.c (working copy) @@ -609